[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [CCAMP] RFC 4207 LMP Erratum



Hi Adrian and Vishwas,
 
I think Vishwas's suggestions are very good  if we want to allow multiple <TRACE> objects in a <TraceMonitor Message>.

However, when we investigate how many <TRACE> objects may be presented on a <TraceMonitor Message>, I think it is natural to include only one <TRACE> ojbect in <TraceMonitor Message>,  because the following format of <TraceMonitor Message> makes the programmer to do that without deep consideration.
  <TraceMonitor Message> ::= <Common Header> <MESSAGE_ID>
                                                     <LOCAL_INTERFACE_ID> <TRACE>
Therefore, I think it depends on how to correct the descriptions of RFC4207 to decide how many <TRACE>objects should be presented in the <TraceMonitor Message>.
 
When we look at other messages, I think <Test> message should also be taken into accout, besides <xxTracexx> messages.
 
(1)Test Message: I think only one <TRACE> object should be included.
 
<Test Message> ::=<Common Header> <LOCAL_INTERFACE_ID>
                                  <VERIFY_ID> <TRACE>
(2) TraceMonitor Message: discussed above.
 
If there are multiple <TRACE> objects in this message, and then there are multiple "Trace Messages". In some cases, if all the Trace objects are processed correctly, we can understand that there is no Trace Mismatch. But if some of them are correct and some of them are incorrect, how to handle that? (i.e.,  in the latter case,  for a specific Local_Interface_ID, is it mismatched  or matched?  Obviously, we must  change <TraceMonitorNack>  or extend ERROR_CODE. ) 
  :-)~~~
 
 
(3) TraceReport Message: I think only one <TRACE> should be included in the <TraceReport Message>.
 
If we look at the format of <TraceReq Message> and <TraceReport Message>, we can get the answer.
<TraceReq Message> ::= <Common Header> <MESSAGE_ID>
                                             <LOCAL_INTERFACE_ID> <TRACE_REQ>
<TraceReport Message> ::= <Common Header> <MESSAGE_ID_ACK> <TRACE>
   The TraceReport message (Message Type 27) is sent over the control channel after receiving a TraceReq message.
 
(4) InsertTrace Message: I think only one <TRACE> should be included in the <InsertTrace Message> for the following description.
 
The InsertTrace message (Message Type 29) is sent over the control channel and is used to request a remote node to send "a specific trace message" over a data link (this assumes that the remote knows the mapping between the local and remote interface_Ids before fulfilling such request).
 
 
Thanks
 
Fatai
 
Advanced Technology Department
Wireline Networking Business Unit
Huawei Technologies Co., LTD.
Huawei Base, Bantian, Longgang,
Shenzhen 518129 P.R.China
Tel: +86-755-28972912
Fax: +86-755-28972935
----- Original Message -----
Sent: Monday, November 02, 2009 12:16 PM
Subject: Re: [CCAMP] RFC 4207 LMP Erratum

Modifying the best BNF format a bit.

> Hi Adrian,
>
> I think the BNF format below is ambiguous as it does not tell which
> part is repeatetive (only trace or others too):
>
> <TraceMonitor Message> ::= <Common Header> <MESSAGE_ID>
> <LOCAL_INTERFACE_ID>
> <TRACE> ...
>
> If we want to be consistent with the other messages in the RFC it should be:
>
> <TraceMonitor Message> ::= <Common Header> <MESSAGE_ID>
> <LOCAL_INTERFACE_ID>
> <TRACE> [<TRACE> ...]
>
>
> However the best representation would be:
>
> <TraceMonitor Message> ::= <Common Header> <MESSAGE_ID>
> <LOCAL_INTERFACE_ID> <TRACE_LIST>
>
> <TRACE_LIST> ::= <TRACE> [<TRACE_LIST> | <empty> ]
>
> Thanks,
> Vishwas
>
> On Fri, Oct 30, 2009 at 2:45 AM, Adrian Farrel <Adrian.Farrel at huawei.com> wrote:
>> Hi CCAMP,
>>
>> Can I please have your help in resolving an Erratum raised against the LMP
>> spec RFC 4207.
>>
>> You can see the full details at
>> http://www.rfc-editor.org/errata_search.php?rfc=4207&eid=167
>>
>> The essence of the question is: how many <TRACE> objects may be present on a
>> <TraceMonitor Message>?
>>
>> The BNF in section 4.1.1 currently implies just one. But the text could be
>> interpreted to man more than one.
>>
>>
>> While you are looking at this, can you say whether any other messages should
>> also allow more than one <TRACE> object?
>>
>> Many thanks,
>> Adrian
>> _______________________________________________
>> CCAMP mailing list
>> CCAMP at ietf.org
>> https://www.ietf.org/mailman/listinfo/ccamp
>>
>
_______________________________________________
CCAMP mailing list
CCAMP at ietf.org
https://www.ietf.org/mailman/listinfo/ccamp